public List<Match> match(List<TextPosition> textPositions, final Pattern pattern) {
StringBuilder sb = new StringBuilder(textPositions.size() * 2);
for (TextPosition textPosition : textPositions) {
if (textPosition != null) sb.append(textPosition.getCharacter());
}
return match(textPositions, sb.toString(), pattern);
}
private String getFont(List<TextPosition> word) {
String baseFont = "";
for (TextPosition position : word) {
baseFont = position.getFont().getBaseFont();
}
return baseFont;
}
private float[] getLineStartEnd(List<TextPosition> line) {
TextPosition first = line.get(0);
TextPosition last = line.get(line.size() - 1);
float[] borderArr = new float[2];
borderArr[0] = first.getXDirAdj();
borderArr[1] = last.getXDirAdj();
return borderArr;
}
@Override
protected void writeCharacters(final TextPosition aText) throws IOException {
if (log.isTraceEnabled()) {
log.trace("[" + aText.getCharacter() + "]");
}
if (regionText == null) {
throw new IllegalStateException("No region started");
}
regionText.append(aText.getCharacter());
}
private void getTheCharacters(List<TextPosition> word) {
for (int i = 0; i < word.size(); i++) {
TextPosition cha = word.get(i);
List<TextPosition> sublist = word.subList(i, i);
db.insertWordCharacter(
rf.convertToCorrectForm(cha),
cha.getXDirAdj(),
cha.getXDirAdj(),
lastInsertedWordId,
rf.isBold(sublist),
rf.isItalic(sublist),
getFont(sublist));
}
}
private void getTheWords(List<TextPosition> line) {
float wordStart = 0;
float wordEnd = 0;
boolean isNextStart = false;
StringBuilder builder = new StringBuilder();
int startIndex = 0;
for (int i = 0; i < line.size(); i++) {
TextPosition pos = line.get(i);
if (i == 0) {
wordStart = pos.getXDirAdj();
builder.delete(0, builder.length());
startIndex = i;
}
if (isNextStart == true) {
isNextStart = false;
wordStart = pos.getXDirAdj();
builder.delete(0, builder.length());
startIndex = i;
}
if (rf.convertToCorrectForm(pos).trim().equals("") || i == (line.size() - 1)) {
if (i == (line.size() - 1)) {
builder.append(rf.convertToCorrectForm(pos));
i++;
}
isNextStart = true;
wordEnd = pos.getXDirAdj();
List<TextPosition> word = line.subList(startIndex, i);
lastInsertedWordId =
db.insertWord(
builder.toString(),
wordStart,
wordEnd,
lastInsertedWorkLineId,
rf.isBold(word),
rf.isItalic(word),
getFont(word));
getTheCharacters(word);
}
builder.append(rf.convertToCorrectForm(pos));
}
}
/**
* A method provided as an event interface to allow a subclass to perform some specific
* functionality when text needs to be processed
*
* @param text The text to be processed
*/
protected void processTextPosition(TextPosition text) {
System.out.println(
"String["
+ text.getXDirAdj()
+ ","
+ text.getYDirAdj()
+ " fs="
+ text.getFontSize()
+ " xscale="
+ text.getXScale()
+ " height="
+ text.getHeightDir()
+ " space="
+ text.getWidthOfSpace()
+ " width="
+ text.getWidthDirAdj()
+ "]"
+ text.getCharacter());
}
/**
* Computes a series of bounding boxes (PDRectangle) from a list of TextPositions. It will create
* a new bounding box if the vertical tolerance is exceeded
*
* @param positions
* @throws IOException
*/
public List<PDRectangle> getTextBoundingBoxes(final List<TextPosition> positions) {
final List<PDRectangle> boundingBoxes = new ArrayList<PDRectangle>();
float lowerLeftX = -1, lowerLeftY = -1, upperRightX = -1, upperRightY = -1;
boolean first = true;
for (int i = 0; i < positions.size(); i++) {
final TextPosition position = positions.get(i);
if (position == null) {
continue;
}
final Matrix textPos = position.getTextPos();
final float height = position.getHeight() * getHeightModifier();
if (first) {
lowerLeftX = textPos.getXPosition();
upperRightX = lowerLeftX + position.getWidth();
lowerLeftY = textPos.getYPosition();
upperRightY = lowerLeftY + height;
first = false;
continue;
}
// we are still on the same line
if (Math.abs(textPos.getYPosition() - lowerLeftY) <= getVerticalTolerance()) {
upperRightX = textPos.getXPosition() + position.getWidth();
upperRightY = textPos.getYPosition() + height;
} else {
final PDRectangle boundingBox =
boundingBox(lowerLeftX, lowerLeftY, upperRightX, upperRightY);
boundingBoxes.add(boundingBox);
// new line
lowerLeftX = textPos.getXPosition();
upperRightX = lowerLeftX + position.getWidth();
lowerLeftY = textPos.getYPosition();
upperRightY = lowerLeftY + height;
}
}
if (!(lowerLeftX == -1 && lowerLeftY == -1 && upperRightX == -1 && upperRightY == -1)) {
final PDRectangle boundingBox = boundingBox(lowerLeftX, lowerLeftY, upperRightX, upperRightY);
boundingBoxes.add(boundingBox);
}
return boundingBoxes;
}
/**
* Write the string in TextPosition to the text cache.
*
* @param text The text to write to the stream.
*/
@Override
protected void writeCharacters(final TextPosition text) {
final String character = text.getCharacter();
textCache.append(character, text);
}
